Many products are produced in successive builds or versions, with each build or version comprising modifications or improvements to the previous build. Testing of such built products commonly involves applying a set of one or more test cases to the product. The quality of the product can then be measured by the degree of the products success against the test cases. However, due to practical constraints, such as time and personnel availability, some testing may not be performed or may be abandoned, while other testing may not be fully effective or reliable. As a result, consistently measuring the quality of the product between subsequent builds is difficult.